Poetry Translating as Expert Action Processes, priorities and networks

 Poetry Translating as Expert Action

to a poem like Krik roughly as I did – especially if my working practices do not

change radically between the two poems.

This cannot be assumed with the two areas of inter-translator variability iden-

tified in Chapter 5. With willingness to consider creative transformations, on Toen

wij I shared with Hugo the moderately-creative mid-point of a ‘literal ↔ radically-

creative’ continuum. Here, only other moderately-creative translators might react

to a poem like Krik roughly as I did. With time management across drafts, how-

ever, my and the others’ Toen wij behaviour differed significantly: hence any com-

parisons with Krik here would only apply to my mid-loaded drafting pattern.

Moreover, Chapter 4 has indicated that not all translators are willing to tackle

rhymed source poems; and of those who are, only some attempt rhymed target

versions, as I do. Any findings relating to rhyme, therefore, are potentially gener-

alizable only to this last category of translators.

6.3 Findings: Toen wij and Krik compared

6.3.1 Task time

My overall translating time for Drafts 1–3 was 2h 41m with Toen wij, and 6h 36m

with Krik. However, Krik is a longer poem. To allow for this, firstly ‘common lexical

items’ (cli) were counted for each source poem. These are items which are obligato-

rily lexicalised in both languages: nouns, main verbs (including the copula am/is/

are), adjectives, participles, adverbs and conjunctions^3. Toen wij had 55, and Krik 98.

Then tape-units per common lexical item were calculated: 38.3 (3m) for Toen wij,

and 52.8 (4m) for Krik. This difference is not statistically significant^4. Krik, therefore,

was not significantly more difficult to translate in time terms than Toen wij.

6.3.2 Drafts and versions

The reason for this lack of significance is shown in Figure 46. I spend less time per

common lexical item in Toen wij than Krik during Draft 1 (8.3 vs. 12.0 tu/cli re-

spectively), and far less during Draft 3 (5.9 vs. 16.8), but times for Draft 2 are the

same (24.1 vs. 24.0). Moreover, Draft 2 is the most intensive draft in both poems,

giving Krik the same mid-loaded pattern as reported in Chapter 5 for Toen wij^5.

3. Simply counting words would not have been valid: ‘of ’, for instance, is a word in Dutch, but
   an adjective/noun suffix in BCS.


4. Paired t-test (tu/cli per draft) t 1.50, 2-tailed p 0.27.


5. Levene’s test for variances: F 1.54, p 0.28, not significant.